Reactions of c i ~ - C r ( C O , ) ( o x ) ~ ~ -
Inorganic Chemistry, Vol. 17, No. 5, 1978 1173 Contribution from the Institute for Physical Chemistry, University of Frankfurt, D6000 Frankfurt a m Main-1, West Germany
Kinetics and Mechanism of Aquation and Formation Reactions of cis-Carbonatobis(oxalato)chromate(III) Ion in Aqueous Solution’ D. A. PALMER, T. P. DASGUPTA,, and H. KELM*
Received September 27, I977 The title compound has been prepared and characterized for the first time. The kinetics of the aquation and formation reactions of this carbonato complex were studied by stopped-flow spectrophotometry. The acid catalysis reactions were studied in the range of 2 > [H’] > 0.01 M and 10 < T < 25 O C at an ionic strength of 2 M (NaC10,). Only the ring-opening reaction of Cr(C0,)(0x),~- is observed, k = 2.04 M-’ s-’ at 25 OC, as subsequent decarboxylation is too fast to measure even at 2 M hydrogen ion concentration. Carbon dioxide uptake by cis-diaquobis(oxalato)chromate(III) was investigated over the ranges 7.01 < pH < 9.34 and 10 < T < 25 O C at an ionic strength of 0.5 M (NaCI). The major reactant species Subsequently, in the pH range studied was ~ i s - C r ( o x ) ~ ( O H ) ( 0 Hwhich ~ ) ~ - takes up C02 to form ~is-Cr(ox)~(HC0~)(0H~)~-. slower ring closure of the latter species to form the carbonato chelate is observed. The results are compared to those of cobalt complexes previously studied and the possible mechanisms are discussed.
Introduction It is well documented3 that the acid-catalyzed aquation of chromium-ammine complexes, viz., CrN jX2+ or CrN4X2+, where N represents an ammine group and X is the monovalent replaceable ligand, is generally complicated by concurrent Cr-N bond fission. However, if the ammine groups are replaced by 0-donating ligands, e.g., oxalate or malonate, substitution of the relatively labile group X- takes place without any side reaction^.^ In continuing our studies on carbonato complexes of various transition-metal ions*-’ we have recently started work on carbonato complexes of chromium(II1). For the reason stated above we have failed to prepare carbonatoamminechromium(II1) complexes. W e found that the Cr-N bonds break rapidly before any measurable association takes place* between the hydroxide of hydroxoamminechromium(111) and carbon dioxide. Apparently this problem does not arise when the ammine ligands are replaced by oxalate groups. In this paper we report the preparation and characterization of the potassium salt of carbonatobis(oxalato)chromate(III) (the oxalato ligand will be hereafter represented by ox) and the results of a detailed kinetic investigation of the aquation and formation of this complex. Experimental Section The cis form of potassium diaquobis(oxalato)chromate(III) dihydrate was prepared by the method of Werner.Io Its composition was confirmed by microanalysis, and the visible-UV spectrum agreed closely with that given in the literature.” Potassium carbonatobis(oxalato)chromate(III) trihydrate was prepared by gradually adding solid K2C03 (ca. 40 mg in 10 cm3 of water) to a stirred solution containing 1 g of ~ i s - K [ C r ( o x ) ~ (OHJ2].2H20 in 10 cm3 of water until a pH of ca. 8.5 was attained. After the solution was stirred for 10 min, it was cooled in an ice bath and ethanol was added until a green precipitate was formed. The crystals were filtered off and thoroughly washed with ethanol, followed by ether. The purity of the complex was checked by microanalysis.I2 K, 25.8; C, 13.07; H , Anal. Calcd for K3[Cr(C03)(C204)2].3H20: 1.32;Cr,11.4. Found: K,26.1;C,13.07;H, 1.32;Cr,11.0. Allother chemicals used were of reagent grade, and deionized, doubly distilled water was used in preparing all solutions. The visible and UV spectra of the various complex species of significance to this research are shown in Figure 1. These were obtained at room temperature using a Zeiss DMRlO recording spectrophotometer. The infrared spectrum of K3[Cr(C03)(ox)2].3H20shows a very weak band at 1025 cm-’ which compares with the reported13 C-0 stretching frequencies at 1593, 1030, and 765 cm-’ for a series of carbonatocobalt(II1) complexes. However the remaining C-0 frequencies of the carbonate are undoubtedly masked by the strong bands of the oxalate ligands. Kinetics. The acid hydrolysis of Cr(C03)(ox)2- was investigated using an Aminco stopped-flow spectrophotometer equipped with a Dasar data storage system. The reaction was followed at 300 nm where
0020-1669/78/1317-1173$01.00/0
the absolute absorbance difference between the carbonato complex and the aquated product is maximal (see Figure 1). The hydrogen ion concentration was regulated with H C I 0 4 and the ionic strength was kept at 2 M using sodium perchlorate. The kinetics of C 0 2 uptake by diaquobis(oxalato)chromate(III) were studied by the “acidification technique” using the abovementioned stopped-flow equipment. The reaction was monitored at 263 nm, where there is a maximum difference in molar absorptivities between the hydroxoaquo or dihydroxo species and the carbonato species. A Tris base-Tris hydrochloride bufferI4 was used to control the pH of the reaction mixture. The ionic strength of these solutions was maintained at 0.5 M (NaC1). It should be noted that NaC10, interferes with the function of the buffer. All pH measurements were made using either a Beckman Research Model or a Radiometer 64 pH meter. Complex concentrations were maintained at lo-’ M throughout this study. The absorbance-time data were transferred to computer punchtapes by means of a teletype connected to the Dasar system, and the rate constants were then computed using a standard least-squares program. pK Determination. The acid dissociation constants K’’ and KI2 of c i s - C r ( ~ x ) ~ ( O H were ~ ) ~ - determined at 10, 15, 20, and 25 OC by M) with 0.1 M N a O H titrating a diaquo complex solution (4 X solution. The ionic strength of these solutions was maintained at 0.5 M with NaC1. The K’’ and K‘, values so obtained are recorded in Table IV.
Results and Discussion The pseudo-first-order rate constants for the acid hydrolysis of Cr(C0,)(0x)~~at four temperatures are summarized in Table I, together with their standard deviations. Linear regressional analysis of these datal5 shows that the rate constants are linear with [H+]over the entire range of concentration studied (0.01-2.0 M). Kinetic runs at hydrogen ion concentrations lower than 0.01 M were abandoned because of concurrent hydrolysis of the oxalate ligands and also because of interaction of the buffer with the substrate. These results could be satisfactorily interpreted in terms of the mechanism previously proposed for the acid hydrolysis of various anionic and cationic cobait(II1) carbonato complexes16-is and can be expressed as k
Cr(C0,)(0x),~-
+ H,O 2 C~(OX),(HCO,)(OH)~’
Cr(C03)(ox),3-
+ H,O’
Cr(ox),(HC0,)(OH,)2-
k
f Cr(ox),(HC0,)(OH2)z-
(2)
k
2 Cr(ox),(OH)(OH2)2- + CO,
(3)
K’
Cr(ox),(HC03)(OH,)Z- + H,O 4 C~(OX),(HCO,)(OH)~+ H,O’ Cr(ox),(OH,),-
+ H,O 2 Cr(ox),(OH)(OH,)2- + H,O’
0 1978 American Chemical Society
(4) (5 1
1174 Inorganic Chemistry, Vol. 17, No. 5, 1978
Palmer, Dasgupta, and Kelm
Table I. Pseudo-First-Order Rate Constants for the Acid Hydrolysis of cis-Carbonatobis(oxalato)chromate(III),I = 2.0 M (NaC10,)
s-'
10Zk",,d,
[Htl, M
10 "C
15 "C
2.0 1.8 1.6 1.4 1.2 1.0 0.75 0.50 0.45 0.40 0.30 0.20 0.15 0.10 0.075 0.050 0.025 0.020 0.010
124 i 6 106 ?- 8 86.1 i 1.3 81.8 i 1.8 68.1 i 0.5 49.8 I1.9 41.5 i 0.9 25.3 i 0.5
162 t 4 147 i 5 135 i 3 115 ? 3 96.6 f 5.8 84.8 i 2.6 56.4 c 1.4 46.8 i 5.2
20 "C
25 "C
272 i 9 262 i 11 229 i- 8 195 r 6 146 t 13 125 i 8 92.5 i 4.1 74.4 i 6.4
520 I: 32 445 i 19 438 t 31 385 i 32 256 ?- 9 208 i 5 131 i 4 101 i 2
40.4 c 0.8 18.2 i 0.5 15.0 t 1.6 9.36 i 0.16 8.56 f 0.52 4.96 i 0.08 4.20 i 0.12 1.78 i 0.06 1.41 i 0.03 0.944 i 0.152
19.6 i 0.4 12.5 i 0.2
3.46
i
0.09
1.18 i 0.04 0.856 i. 0.004 0.484 i. 0.024
49.6 A 4.8 34.5 i 1.5 22.6 t 0.2 10.2 t 0.6 7.88 I: 0.56 5.24 i 0.44 2.42 i 0.11 2.21 i 0.08 1.28 i 0.02
52.4 + 0.3 45.2 i 2.4 19.9 i 0.5 18.8 i 0.9 10.2 ?- 2.4 8.80 i- 0.16 4.40 i 0.28 3.99 i 0.06 1.31 i 0.05
Table 11. Rate Parameters for the Acid-Catalyzed Ring Opening of Some Anionic Chelated Cobalt(II1) and Chromium(II1) Carbonato Complexes at 25 "C, I = 2.0 M k,, IV1- '
1.
...,
'
200
300
y
. .....,
"..............
"
LOO WAVELENGTH
500
600
Complex ion
S-,
Co(CO,)(NTA)*a-Co(CO,)(EDDA)P-Co(C0 ,)(EDDA)CO(C0 3 )klY ) 2 Cr(CO,)(ox), 3 -
42 118 2.4 8.0 2.0
AH',
AS',
kcal mol-'
cal deg-' mol-'
16 12 11 22 14.7
-2 -8 -20 +20 -8
Ref 16 17 17 20 This work
r
700
I nm I
Figure 1. Visible and UV absorption spectra of the complexes: c i s - C r ( ~ x ) ~ ( O H , )---, ~ ; ~ i s - C r ( o x ) , ( O H ) ( 0 H , ) ~ ~ ; -, cis-Cr( o x ) ~ ( o H ) ~ ~-,- ; C ~ ( C O , ) ( O X ) ~ ~ ? ..e,
- a
According to this reaction sequence the expected rate expression should be kokd = ko + kl [H'], However the k , values are too small to be observed in the range of acid concentration s t ~ d i e d . Hence '~ the rate expression reduces to kokd= kl[H+]. The k l values, obtained using a "best fit" program, yielded the activation parameters A H t = 14.7 f 0.6 kcal mol-' and AS* = -8 f 2 cal deg-' mol-'. From the mechanism proposed above it is clear that the k&d values should increase linearly with [H'] and then start to level off to a constant value a t high acidity. This constant value should represent the rate constant for decarboxylation of the aquobicarbonato species, k 2 in eq 3. The data in Table I clearly show that even at 2 M [H'] there is no indication of leveling off. This immediately suggests that the rate of decarboxylation of the aquobicarbonato species is much faster than the rate of ring opening. Since decarboxylation occurs by carbon-oxygen bond fission, the reactivity is not expected to be dependent on the nature of the metals5 Assuming a decarboxylation rate constant (k,) for Cr(ox)2(HC03)(OH2)2of ca. 55 s-l l 8 and knowing the ring-opening rate constant ( k , ) is 2.04 M-* sS1, then at 2 M [H'], kobsd = 4.08 s-'. Thus, a t the highest acid concentration studied, k2 is nearly 14 times greater than kl, which justifies our experimental observation. It is of interest to compare the rate parameters for the acid-catalyzed ring-opening process of C r ( C 0 3 )( 0 ~ ) with ~~those of various anionic carbonatocobalt(II1) complexes. These data are shown in Table 11. It is obvious that the overall charge on the complex ion is not a significant factor in determining the rate of reaction. As proposed earlier,'* the
t Figure 2. Oscilloscopic traces for C 0 2 uptake by cis-Cr(ox)2(OH)(OH,),- at p H 8.20 and 25 " C with [CO,] = 0.01 M. The ordinate is in % transmission, while the abscissa is calibrated at (a) 0.1 s/division and (b) 2 s/division.
relative magnitudes of the rate constants appear to depend on the stereochemical requirements of the "inert" ligands. Thus stereochemical reorganization, as well as solvation properties, reflect in the activation parameters. The nature of the central metal atom should not be a prime factor in determining the reactivity, since the rates of aquation of Cr(II1) complexes do not differ very much from those of the Co(II1) complexes, although the substitution reactions of the former complexes are believed to occur by an I, mechanism.21 Stopped-flow traces for the C 0 2 uptake by cis-Cr(ox)2(OH) clearly show the existence of two kinetically distinguishable successive reactions, as can be seen in Figure 2. The initial rapid rate corresponds to the formation of a monodentate carbonato species (Figure 2a). The trace in Figure 2b then represents the relatively slow ring closure of the coordinated carbonate. These results are entirely consistent with previous studies of C02uptake by hydroxoaquocobalt(II1) complex~s.~~~ The C 0 2 uptake study was conducted in the ranges 10 < temperature < 25 "C and 7.01 < pH < 9.34, and the resulting pseudo-first-order rate constants are recorded in Table 111.
Reactions of c i ~ - C r ( C O , ) ( o x ) ~ ~ -
Inorganic Chemistry, Vol. 17, No. 5, 1978 1175
Table 111. Rate Constants‘ for the Carbon Dioxide Uptake by Cr(ox),(OH,),10 “C PH 7.22 7.23 7.25 7.36 7.42 7.58 7.68 7.69 7.77 7.78 7.90 8.58 8.79 9.34
‘[CO,]=
15 “C PH 7.26 7.33 7.36 7.41 7.48 7.49 7.53 7.60 7.65 7.68 7.76 7.90 7.95 8.12
lok’obsd
2.41 I 0.07 2.00 t 0.01 1.57 f 0.06 1.77 f 0.04 2.16 * 0.04 2.47 f 0.07 2.56 f 0.05 2.30 I 0.06 1.67 t 0.03 1.93 I 0.05 1.32 t 0.04 0.87 t 0.03 0.79 t 0.02 0.89 f 0.04
20 “ C
lok‘obsd
2.62 t 0.24 2.67 f 0.14 2.43 I 0.13 2.76 i. 0.06 3.14 I 0.49 2.51 i: 0.24 3.08 f 0.34 2.95 i 0.07 3.55 f 0.48 2.36 f 0.06 2.96 f 0.06 2.78 * 0.10 1.98 * 0.05 2.93 f 0.03
PH 7.01 7.10 7.17 7.25 7.34 7.40 7.43 7.52 7.56 7.60 7.71 7.97 9.18
25 “C
lok’obsd
5.28 3.97 4.43 4.47 4.87 5.10 5.14 4.76 4.98 4.92 4.48 6.63 5.66
0.09 t 0.08 IO.10 f 0.06 t 0.20 1. 0.12 f 0.15 0.20 t 0.09 i 0.09 i 0.15 f 0.20 0.26 t
*
*
PH 7.04 7.09 7.18 7.31 7.40 7.46 7.78 7.85 7.92 8.03 8.04 8.20 8.22 8.64 8.69 8.70
lok’obsd
5.72 ?; 0.39 7.77 f 0.04 6.90 f 0.31 4.82 k 0.09 5.71 k 0.08 7.97 t 0.17 8.58 t 0.25 4.49 I 0.09 3.96 t 0.05 7.08 f 0.24 9.47 t 0.27 6.07 f 0.19 3.63 t 0.60 9.0 i 0.9 6.26 f 0.52 11.2 I 0.4
0.01 M ; I = 0.5 M, NaCl. kIobsd has the units s-,.
Table IV. Rate Parameters for the Carbon Dioxide Uptake by Cr(ox)z(OHz)2-
‘
T,“C 25 20 15 10
k ’ , , M - l s-l 107 71 50.5 34
PK‘, 6.96 7.00 7.08 7.13
PK’Zb 8.97 9.01 9.12 9.19
1200r_
0”
u 1000-
-2
‘A H I * = 12.1 f 0.3kcalmol-’;AS,*=-8.61
1.2caldeg-‘ mol-’. The errors in pK‘, and p K , are ca. 0.03 and 0.06, respectively.
-2
Scheme I cis-Ce ! O X )
+ +-
2‘ w -
-=Y
(OHZ)
600 -
r 600-
I
-52-
+
N
+-! I
400-
IO8 [PI
These data are interpreted in terms of the reaction sequence in Scheme I. The pK‘, and P K ‘ ~values determined in this study are in reasonable agreement with earlier measurements.22 Within the range of p H used here, the “aquo” complex ion consists of a pH-dependent mixture of all three species: cis-Cr( O X ) , ( O H ~ ) ~c- i, ~ - C r ( o x ) ~ ( 0 H ) ( O H ~and ) ~ - ,c i s - C r ( ~ x ) ~ (OH):-. Thus, the reactions in Scheme I require that for the relatively rapid reactions described by k’l and k’26
Figure 3. Plots according to the rate expression (7) a t various temperatures. Table V. Rates of CO, Uptake by Various Hydroxoaquo Complexes of Cobalt(II1) and Chromium(II1) a t 25 OC, I = 0.5 M
Complex ion
k‘,, M-’ AH*,kcal AS*,cal s-’ mol-’ deg-’ mol-’
Co(NH,),OHZ+ 220 15.3 f 0.9 3.6 Co(tren)(OH,)(OH)Z+ 44 14.7 t 0.1 -1.9 d cis-Co(en),(OH,)(OH)2+ 240 15.3 f 0.9 3.4 - -(total “aquo” complex) = ~’,(C~(OX),(OH)(OH~)~-) cis-Co(cyclam)(OH,)(OH)Z+ 57 14.8 t 1.0 0.8 dt Cr(NH,),0H2+ 7 Cr(ox), (OH,)(OH)2107 12.1 i 0.3 -8.6 k‘2(Cr(ox)2(0H)23-) [COZl (6) ~
{
+
1
which rearranges to give the observed pseudo-first-order rate constant k’obsd: k’obsd =
(k’lK’1 [H’] f k’2Kt1K’2)[C02] [H’]’ K‘, [H’] f K I 1 K f 2
+
+
(7)
Plots of ([H+]’+ K’,[H+] K’lK’2)kbbsd/K’l[C02] VS. [H’] yield straight lines with slopes k’, and intercepts of k’2K’2, as shown in Figure 3. The PI,K’,, and K f 2values a t various
Ref
3.0 f 0.2 ?: 3.2 f 3.5 t
i
9 6 25 7 26 1.2 This work
temperatures are listed in Table IV. As is evident from Figure 3, the intercept a t each temperature is negligibly small. This phenomenon is certainly not due to the inactivity of the C r ( o ~ ) ~ ( o Hspecies, ) ~ ~ - but rather to the large uncertainty in klobsda t higher p H due to the slow concurrent hydrolysis of the oxalate ligands23 and to the more rapid hydrolysis of C 0 2 a t high pH.24 A further contributing factor is the probable similarity between the spectra of cis-Cr(ox)2(0H),3-
1176 Inorganic Chemistry, Vol. 17, No. 5, 1978 Table VI. Rate Constants for the Ring Closure of C~(OX),(HCO,)(OH)~at 25 “ C , I = 0.5 M (NaC1) 7.09 7.17 7.18 7.34 7.40 7.46 7.48 7.56
4.28 t 0.03 2.10 * 0.28 1.27 -?. 0.17 2.70 * 0.06 3.41 i 0.04 2.85 i. 0.11 2.76 * 0.30 2.02 2 0.15
7.60 7.71 7.78 7.91 8.03 8.04 9.18
2.10 * 0.04 2.21 i 0.19 2.28 i 0.19 1.48 i 0 1 2 1.56 i 0.02 1.67 i 0.11 2.70 i 0.07
and c i s - C r ( o ~ ) ~ ( C O(, ) It is now of interest to compare the rate parameters for COz uptake by the more extensively studied cis-aquoammine complexes of ~obalt(II1)~ with the results of the present system. These rate parameter data are summarized in Table V. It is apparent that k’l for cis-Cr(ox)z(OH)(OHz)2-is very similar to the corresponding values for the cobalt(II1) complexes. This similarity also exists between the activation parameters. The entropy of activation for the bis(oxa1ato) complex is slightly more negative than the others which emphasizes that little or no disruption of the solvation shell takes place. The rate constant for COz uptake by Cr(NH,),(OH)2+ has been estimated to be only 7 M-’ s-’ as a lower limit from a study26 of CO,-catalyzed 0 exchange nith this complex. This value is significantly lower than k’’ determined for cis-Cr(ox)2(OH)(OH,)*-, which is surprising as one would not expect the magnitude of the charge on the substrate to affect the rate of such a direct associative reaction involving a neutral COz molecule. It is noteworthy that the formation of a stable Cr(?JH,),CO,+ ion was not established26and the mechanism of exchange is based on the formation of an unstable carbonato complex intermediate. Recently, Dasgupta and Harris’ showed that a reasonable correlation exists between log k for the COz addition to coordinated hydroxide and the pK, of the conjugate aquo acid. The rate of C 0 2 uptake by cis-Cr(ox)z(OH)( 0 H J 2 - satisfactorily fits into that correlation (see Figure 3 in ref 7). However, the log k value for Cr(hH3)s(OH)2twas found to be too small to reasonably fit the linear correlation. The ring closure rate data for c i ~ - C r ( o x ) , ( H C O ~ ) ( 0 H ) ~ are presented in Table VI. These rate constants correspond to the comparatively slow release of oxalate at high pH. Therefore, these results cannot be discussed in detail. Nevertheless, it is interesting to note that the rate of ring closure is similar to that observed for cobalt(II1)-ammine monodentate carbonato complexe~,~ Le., in the range s-l. This alone suggests that ring closure of the carbonatoaquo complexes of chromium(II1) also follow a Sh2(internal) mechanism, as proposed earlier.7 However, an unequivocal
Palmer, Dasgupta, and Kelm assignment of the mechanism for ring closure necessitates additional studies of other chromium(II1) activation parameters. Acknowledgment. The authors gratefully acknowledge the financial assistance of the Deutsche Forschunsgemeinschaft. Registry No. Cr(C0,) OX)^^-, 659 15-84-6; K, [ Cr(C0,)(0x)~]. 65890-26-8; H’, 12408-02-5; c i ~ - C r ( o x ) ~ ( O H15489-30-2; ~)~, K2C03, 584-08-7.
References and Notes A preliminary report of this paper was presented at the ICCC, Hamburg, Germany, 1976. On leave from the Department of Chemistry, University of the West Indies, Mona, Kingston 7, Jamaica. T. P. Dasgupta, MTP Int. Rec. Sci.: Inorg. Chem., Ser. Two,9,76 (1974). (a) T. W. Kauen, Inorg. Chem., 15, 440 (1976); (b) H. Kelm and G. M. Harris, ibid., 6, 1743 (1967); (c) D. Banerjea and M. S. Mohan, J . Inorg. Nucl. Chem., 27, 1643 (1965); (d) K. V . Krishnamurthy and G. M. Harris, J . Phys. Chem.: 64, 346 (1960). (a) Reference 3, p 73; (b) D. A. Palmer and G. M. Harris, Inorg. Ckem., 13, 965 (1974). T. P. Dasgupta and G. M , Harris, J . A m . Chem. Soc., 97, 1733 (1975). T. P. Dasgupta and G. M. Harris, J . A m Chem. Soc., 99,2490 (1977). It has been demonstrated earlier that the aqua-ammine complexes of c ~ b a l t ( I I I ) ~do ~ ’not . ~ react with COz. Similarly, we found that the diaquobis(oxalato)chromate(III) ion in a C02-saturated solution does not form the carbonato complex on long standing. E. Chaffee, T. P. Dasgupta, and G . M. Harris, J . A m . Chem. Soc., 95, 4169 (1973). A. Werner, Justus Liebigs Ann. Chem., 406, 299 (1914). H. Kelm and G . h?. Harris, Inorg. Chem., 6, 706 (1967). I. Beetz, Mikroanalytisches Laboratorium, 8640 Kronach. K. Nakamoto, “Infra-red Spectra of Inorganic Cc-ordination Compounds”, Wiley, New York. N.Y., 1970. p 170. Tris = tris(hydroxymethy1)aminomethane. Due to the scattering of the data at high [H’], a slightly negative intercept was found using the standard least-squares treatment. The values of k l were therefore determined using a best-fit program of the form y = mx.
T. P. Dasgupta and G. 51. Harris, J . A m . Chem. Soc., 93, 91 (1971). T. P. Dasgupta and G. M Harris, Inorg. Chem., 13, 1275 (1974). R. van Eldik, T. P. Dasgupta, and G. M. Harris, Inorg. Chem., 14. 2573 (1975). Usually the difference between the acid-catalyzed and spontaneous ring-opening rate constants is in the range of 3-4 orders of magnitude in the case of cobalt(II1) complexes (see ref 15). Similar differences may be expected for chromium(II1) complexes. T. P. Dasgupta, unpublished results. J. 0. Edwards. F. Monacelli. and G . Ortaggi, Inorg. Chim. Acta, 11, 61 (1974). H. Kelm and G. M. Harris, 2.Phys. Chem. (Frankfurt am Main), 66, 8 (1966). The pK values obtained at 20 ‘C are 7.1 and 9.3. T. 4’. Kallen, Inorg. Chern., 14, 2687 (1975). R. E. Forster, J. 7. Edsall, A. B. Otis. and F. J. W .Roughton, Ed., “CO2, Biochemical and Physiological Aspects”, Kational Acronautics and Space Administration, Washington, D.C., 1969 (obtainable from the U.S. Government Printing Office, Washington, D.C.). J. M. DeJovine, W.K. Wan, and G. M. Harris, to be submitted for publication. J. E. Early and W. Alexander, J . A m . Chem. Soc., 92, 2299 (1970).
